VISUAL OUTDOOR PATH PLANNER FOR ORANGE GROVES BASED ON ENSEMBLES OF NEURAL NETWORKS

Joaquín Torres-Sospedra, Patricio Nebot

Abstract

One of the most important system to deploy for a robot navigating in an outdoor scenario, as can be an orange grove, is the navigation system. In this paper, a path planner in orange groves for an autonomous robotic system is presented. This path planner is based on a previous classification of the image that the robot gets from its visual sensory system. One of the most important technique used to generate accurate classifiers is based on training an ensemble of neural networks. Here, a simple ensemble of neural networks is used to classify images from an orange grove using wavelets features. With the classification image obtained, the most important lines of the land are extracted with the Hough transform. The final path line is determined with these lines. The purpose of this paper is to determine if the ensemble approach can be useful in the procedure to design an accurate path planner for outdoor autonomous robots in orange groves. The published results show that ensembles can be considered for this type of applications.

References

  1. Bishop, C. M. (1995). Neural Networks for Pattern Recognition. Oxford University Press, Inc., New York, NY, USA.
  2. Dietterich, T. G. (2000). Ensemble methods in machine learning. In Kittler, J., editor, Multiple Classifier Systems., number 1857 in LNCS, pages 1-15.
  3. Fernndez-Redondo, M., Hernndez-Espinosa, C., and Torres-Sospedra, J. (2004). Multilayer feedforward ensembles for classification problems. In Pal, N. R., Kasabov, N., Mudi, R. K., Pal, S., and Parui, S. K., editors, Neural Information Processing, 11th International Conference, ICONIP 2004, Calcutta, India, November 22-25, 2004, Proceedings, volume 3316 of Lecture Notes in Computer Science, pages 744-749. Springer. ISBN: 3-540-23931-6.
  4. Kuncheva, L. I. (2004). Combining Pattern Classifiers: Methods and Algorithms. Wiley-Interscience.
  5. Oza, N. C. (2003). Boosting with averaged weight vectors. In Multiple Classifier Systems, volume 2709 of LNCS, pages 15-24. Springer.
  6. Raviv, Y. and Intratorr, N. (1996). Bootstrapping with noise: An effective regularization technique. Connection Science, Special issue on Combining Estimators, 8:356- 372.
  7. Sung, G.-Y., Kwak, D.-M., and Lyou, J. (2010). Neural network based terrain classification using wavelet features. Journal of Intelligent and Robotic Systems, 59(3-4):269-281.
  8. Torres-Sospedra, J., Hernndez-Espinosa, C., and FernndezRedondo, M. (2005). New results on ensembles of multilayer feedforward. In Artificial Neural Networks: Formal Models and Their Applications., volume 3697 of Lecture Notes in Computer Science, pages 139- 144. Springer.
  9. Tumer, K. and Ghosh, J. (1996). Error correlation and error reduction in ensemble classifiers. Connection Science, 8(3-4):385-403.
  10. Verikas, A., Lipnickas, A., Malmqvist, K., Bacauskiene, M., and Gelzinis, A. (1999). Soft combination of neural classifiers: A comparative study. Pattern Recognition Letters, 20(4):429-444.
Download


Paper Citation


in Harvard Style

Torres-Sospedra J. and Nebot P. (2011). VISUAL OUTDOOR PATH PLANNER FOR ORANGE GROVES BASED ON ENSEMBLES OF NEURAL NETWORKS . In Proceedings of the 8th International Conference on Informatics in Control, Automation and Robotics - Volume 2: ICINCO, ISBN 978-989-8425-75-1, pages 223-228. DOI: 10.5220/0003537602230228


in Bibtex Style

@conference{icinco11,
author={Joaquín Torres-Sospedra and Patricio Nebot},
title={VISUAL OUTDOOR PATH PLANNER FOR ORANGE GROVES BASED ON ENSEMBLES OF NEURAL NETWORKS},
booktitle={Proceedings of the 8th International Conference on Informatics in Control, Automation and Robotics - Volume 2: ICINCO,},
year={2011},
pages={223-228},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0003537602230228},
isbn={978-989-8425-75-1},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 8th International Conference on Informatics in Control, Automation and Robotics - Volume 2: ICINCO,
TI - VISUAL OUTDOOR PATH PLANNER FOR ORANGE GROVES BASED ON ENSEMBLES OF NEURAL NETWORKS
SN - 978-989-8425-75-1
AU - Torres-Sospedra J.
AU - Nebot P.
PY - 2011
SP - 223
EP - 228
DO - 10.5220/0003537602230228